Origin of the Angiospcrmous Flora of South Africa. 361 



Eurasian types into Africa and there commenced in numerous 

 regions a more diverse development than in its original home. 



The absence in South Africa of so many of those elements which 

 have found their way to the mountains of the most southern parts 

 of America and Australasia, makes the presence of others all the 

 more remarkable. The presence of many of these find their explana- 

 tion in the distribution of the plants of the higher mountains of 

 Africa (compare Engler, 1892, p. 85 and following). Others, how- 

 ever, are at present quite isolated. Thus Fedde (1905, p. 42) 

 remarks as follows on the Papaveracecd : " The rather isolated occur- 

 rence of PapaveracecB at the Cape has many analogies in other 

 relations between the Cape Flora and the Mediterranean Region. 

 The occurrence of Fumaria abyssinica in Abyssinia and a new 

 species of Fumaria on the Kilimandjaro does not clear up this 

 matter, since in South Africa only the genera Papaver, Corydalis 

 and Trigonocapnos occur, but not Fumaria" (except F. officinalis, 

 L., as an introduced weed). " Perhaps we may yet find repre- 

 sentatives of the last three genera on the mountains of Eastern 

 Africa." 



Fedde also remarks upon the extraordinary distribution of our 

 only species of Papaver, P. aculeatum, Thunb., which occurs only 

 with us and in East and South Australia, but from what I know of 

 this species in South Africa, I venture to say that there can be 

 little doubt that it is not a native species. 



However, it was not my object in this essay to aim at an 

 exhaustive treatment of the subject of the origin of the Angio- 

 spermous Flora of South Africa, nor would I have been able to 

 accomplish this object. If I have succeeded in showing even dimly 

 in what channels the elements comprising the Angiospermous Flora of 

 South Africa have moved since Cretaceous times I shall feel amply 

 rewarded. It can scarcely be hoped that we shall ever get a per- 

 fectly clear view, for numerous types, of which we have no concep- 

 tion, must have died out since Angiospermous plants first made 

 their appearance on this globe ; further, we have no absolute 

 standard, and are not likely to get one, by which we can with 

 ■certainty fix the relative ages of the majority of Angiospermous 

 types ; further still, we have no absolute means by which we can 

 determine the direction in which the current of plant-development 

 has moved, and, lastly, while we may dimly perceive the main 

 channels of such development, there must have been numerous 

 -side- and cross-channels of various widths and various lengths. 

 These must have frequently influenced the final result. At the 

 same time we must acknowledge that Zoo-geography, assisted by 



24 



